EP3625627B1 - Sum streams for actual conditions and control signals of a distributed control system - Google Patents

Description

Control systems are known in various configurations. Historically, the control systems were initially set up with independent communication systems, for example a separate bus of a programmable logic controller. Further developments then led to fieldbus systems in which the sensors and actuators are connected to the common central unit via a fieldbus - for example the PROFIBUS.

More recently, field buses that work on an Ethernet basis have also been used. Such a field bus is a typical example of an open communication network.

An open communication network is a communication network in which each component connected to the communication network feeds data into the communication network according to its own specifications and also reads data transmitted via the communication network. In contrast, the connected components do not know which other components are connected to the communication network. In particular, additional components that have nothing to do with the control can thus also be connected to an open communication network that is used within the scope of the present invention. If - purely by way of example - the central unit, the sensors and the actuators are connected to the communication network as components, additional components can be present that communicate with each other, but neither with the central unit nor the sensors nor the actuators, or with at least one of these components communicate, but not within the tax process. Furthermore, the connected components - at least as a rule - do not know the topology of the open communication network.

In the process industry and industrial automation, the industrial technical processes to be controlled are usually relatively sluggish. Cycle times - "time window" in the terminology of the present invention - can be in the range of several milliseconds, often even in the range above 50 ms or 100 ms. As a rule, an Ethernet-based communication network can also be used for such cycle times.

From the DE 10 2015 213 845 A1 a control method for a technical process in the form of controlling components within a vehicle is known. As part of this control method, sensors of a control system record cyclical actual states of the industrial technical process and transmission of the actual states via a first open communication network to a common central unit of the control system. The common central unit, taking into account the actual states transmitted to it, cyclically determines control signals for the industrial technical process and transmits the control signals to actuators of the control system via a second open communication network. The actuators act cyclically on the technical process in accordance with the control signals transmitted to them. The sensors transmit the actual status cyclically. However, reception by the common central unit within predetermined times cannot be guaranteed with certainty. The same applies to the reverse communication between the common central unit and the actuators. In the case of citation D1, different times can be specified at which the individual sensors transmit their actual states to the central unit.

In the context of the transmission of audiovisual data streams, specially protected connections, so-called streams, are known in the prior art for open communication networks. A stream is a connection between a first component and a second component, both of which are connected to the open communication network. In the case of a stream, it is ensured that the data fed into the first component - the so-called talker - arrive at the second component - the so-called listener - with a maximum delay known in advance. The maximum delay depends on the line sections and the individual nodes between successive line sections from the first to the second component. The maximum delay is set when the stream is set up.

The measures known for the audiovisual transmission of data can in principle also be used in an industrial environment. However, there is an important difference between audiovisual data and industrial data. at Audiovisual data usually have to transmit large amounts of data from exactly one talker to exactly one listener. In the case of industrial applications, on the other hand, the amount of data to be transmitted is already considerably smaller. Usually only a few bytes, a maximum of a few kbytes, have to be transferred per cycle. Furthermore, in industrial applications, not only two components are involved in communication, but many components. In particular, each sensor is a talker for the transmission of the actual states, while the central unit is a listener for this. Conversely, the central unit is a talker for transmitting the control signals, while each actuator is a listener.

If the measures known from the audiovisual transmission of data were to be adopted directly and immediately, a large number of streams would have to be set up, namely a separate stream for each sensor and each actuator, with the respective sensor being the talker and the central unit being the listener and with the actuators the central unit is the talker and each actuator is a listener. Each stream occupies considerable resources at each node of the communication network through which the stream is routed. This approach therefore proves to be possible in principle, but not practicable.

In the prior art, the data from control systems (i.e. the actual states detected by the sensors and the control signals to be transmitted to the actuators or, in general, the process image) are therefore transmitted without any further protection. A high priority is selected for the transmission and a rough estimate of the bandwidth required for this is carried out.

As far as the control signals to be transmitted from the common central unit to the actuators are concerned, the prior art procedure is relatively problem-free. In particular, it can be ensured in a relatively simple manner on the part of the common central unit that the Control signals for the actuators are fed sequentially one after the other into the open communication network. However, the sensors feed their current states into the open communication network in an uncoordinated manner. "Beating" can therefore occur in the communication network. A beating occurs when several sensors feed their actual states into the communication network in such a way that the central unit or one of the nodes is briefly overloaded. Due to the overload, the central unit or in the affected node point cannot buffer the data of one of the relevant sensors, so that there is a loss of data. Such a loss of data usually leads to a disruption of the control process as such.

Other events can also influence the control process. For example, the latency of individual nodes can change due to a different communication that is not related to the control process as such. This can also lead to a brief overload of a node.

Such problems could be avoided by setting up streams between the components involved (sensors, common central unit, actuators). However, setting up streams between only two components at a time, as is the case by default, is impractical.

The object of the present invention is to create possibilities for realizing protected communication between the components involved with reasonable effort.

The object is achieved by a control method with the features of claim 1. Advantageous refinements of the control method are the subject matter of the dependent claims 2 to 5.

• dass die Sensoren die Istzustände über eine den Sensoren gemeinsame erste geschützte Verbindung eines ersten offenen Kommunikationsnetzes an die gemeinsame Zentraleinheit übermitteln,
• dass die gemeinsame Zentraleinheit die Istzustände über eine den Aktoren gemeinsame zweite geschützte Verbindung eines zweiten offenen Kommunikationsnetzes an die Aktoren übermittelt und
• dass jedem Sensor innerhalb des Zeitfensters ein jeweiliger senderseitiger Teilbereich zugeordnet ist, innerhalb dessen der jeweilige Sensor die von ihm erfassten Istzustände dem ersten offenen Kommunikationsnetz zuführt,
• dass die übermittelten Istzustände innerhalb eines mit dem jeweiligen senderseitigen Teilbereich korrespondierenden jeweiligen empfängerseitigen Teilbereichs des Zeitfensters bei der Zentraleinheit eintreffen und
• dass die senderseitigen Teilbereiche der Sensoren derart bestimmt sind, dass die empfängerseitigen Teilbereiche zueinander disjunkt sind.

According to the invention, a control method of the type mentioned is designed in that

• that the sensors transmit the actual states via a first protected connection common to the sensors of a first open communication network to the common central unit,
• that the common central unit transmits the actual states to the actuators via a second protected connection of a second open communication network that is common to the actuators and
• that each sensor is assigned a respective transmitter-side sub-area within the time window, within which the respective sensor feeds the actual states it has detected to the first open communication network,
• that the transmitted actual states arrive at the central unit within a respective receiver-side sub-area of the time window that corresponds to the respective transmitter-side sub-area, and
• that the transmitter-side partial areas of the sensors are determined in such a way that the receiver-side partial areas are disjoint from one another.

According to the invention, a total of only two protected connections are set up, namely on the one hand the first protected connection between the sensors and the common central unit and on the other hand the second protected connection between the common central unit and the actuators.

The communication from the common central unit to the actuators via the common second protected connection is relatively problem-free. In contrast to the usual procedure for audiovisual streams, there are indeed more than two participants. Of the participants, however, only one participant, namely the common central unit, feeds data into the second protected connection. The common central unit can therefore output the control signals to the actuators one after the other. Each actuator can listen in and out via the second shared protected connection transmitted data filter out the control signals intended for him.

With regard to the communication from the sensors to the common central unit, however, as in the prior art, none of the sensors "know" that the other sensors also transmit the actual states recorded by them to the central unit via the common first protected connection. In order to ensure that there are no temporal conflicts when transmitting the actual states to the common central unit, each participating sensor is assigned the respective transmitter-side sub-area within the time window, within which the respective sensor feeds the actual states it has detected to the first open communication network . This ensures that the transmitted actual states arrive sequentially one after the other at the central unit.

• dass für die Sensoren der jeweilige senderseitige Teilbereich innerhalb des Zeitfensters zunächst vorläufig angesetzt wird,
• dass sodann für jeden senderseitigen Teilbereich ermittelt wird, welcher empfängerseitige Teilbereich auf Seiten der gemeinsamen Zentraleinheit mit dem jeweiligen senderseitigen Teilbereich korrespondiert, und
• dass schließlich die senderseitigen Teilbereiche innerhalb des Zeitfensters derart verschoben werden, dass die empfängerseitigen Teilbereiche zueinander disjunkt sind.

In order to correctly determine the sender-side sub-areas for the data transmission via the first protected connection, it has proven to be advantageous

• that the respective transmitter-side sub-area within the time window is initially provisionally set for the sensors,
• that it is then determined for each transmitter-side sub-area which receiver-side sub-area on the side of the common central unit corresponds to the respective transmitter-side sub-area, and
• that finally the sub-areas on the transmitter side are shifted within the time window in such a way that the sub-areas on the receiver side are disjoint from one another.

As a rule, a certain safety distance between the receiver-side subregions is also maintained. However, this is not absolutely necessary.

The communication from the respective sensor to the common central unit takes place via a respective sequence of sequential successive line sections of the first open communication network. In this case, line sections of the respective sequence that are directly adjacent to one another are connected to one another via a respective node (a so-called bridge). Each node forwards the actual states transmitted by the respective sensor with a respective delay time. Each node also ensures for each protected connection in which it is involved that the delay time with which it forwards the data it has received lies between a minimum value predetermined for the respective node and a maximum value predetermined for the respective node. This applies to every protected connection and thus also to the first protected connection of the first communication network. Starting from the transmitter-side sub-area assigned to the respective sensor, it is thus possible, by adding up the predetermined minimum values of the nodes, to determine an earliest possible point in time at which the corresponding actual states can arrive at the common central unit at the earliest. In an analogous manner, starting from the transmitter-side sub-area assigned to the respective sensor, it is possible by adding up the predetermined maximum values of the nodes to determine a latest possible point in time at which the corresponding actual states can arrive at the common central unit at the latest. The respective receiver-side sub-area is thus determined by the earliest possible point in time and the latest possible point in time. If the sender-side sub-area is later shifted, the corresponding receiver-side sub-area shifts 1: 1 with this shift. If, for example, the sub-area on the transmitter side is brought forward by 50 microseconds, the corresponding sub-area on the receiver side is also shifted forward by 50 microseconds.

The maximum values of the nodes must be given explicitly. The minimum values can be given explicitly. Alternatively, it is possible for the predetermined minimum values to be set to zero.

As a rule, the first open communication network and the second open communication network have at least one common line section via which both the transmission of the actual states to the common central unit and the transmission of the control signals to the actuators take place. In particular, there is often only a single Ethernet as a (common) communication network, to which the common central unit is connected once (1x), so that all communications between the common central unit with the sensors and the actuators must take place via this connection.

• dass die Sensoren und die gemeinsame Zentraleinheit über ein erstes offenes Kommunikationsnetz miteinander verbunden sind,
• dass die gemeinsame Zentraleinheit und die Aktoren über ein zweites offenes Kommunikationsnetz miteinander verbunden sind und
• dass die Sensoren, die gemeinsame Zentraleinheit, die Aktoren, das erste offene Kommunikationsnetz und das zweite offene Kommunikationsnetz derart ausgebildet sind, dass sie im Betrieb gemäß dem Steuerverfahren miteinander zusammenwirken.

The object is also achieved by a control system having the features of claim 6. According to the invention, a control system of the type mentioned is designed in that

• that the sensors and the common central unit are connected to one another via a first open communication network,
• that the common central unit and the actuators are connected to one another via a second open communication network and
• that the sensors, the common central unit, the actuators, the first open communication network and the second open communication network are designed in such a way that they interact with one another during operation in accordance with the control method.

FIG 1

ein Steuerungssystem und einen industriellen technischen Prozess,

FIG 2

ein Zeitdiagramm,

FIG 3

eine geschützte Verbindung,

FIG 4

ein Datenpaket,

FIG 5

eine weitere geschützte Verbindung,

FIG 6

ein weiteres Datenpaket,

FIG 7 bis 11

Zeitdiagramme, und

FIG 12

ein Steuerungssystem.

The properties, features and advantages of this invention described above and the manner in which they are achieved will become clearer and more clearly understood in connection with the following description of the exemplary embodiments, which are explained in more detail in conjunction with the drawings. Here show in a schematic representation:

FIG 1

a control system and an industrial technical process,

FIG 2

a timing diagram,

FIG 3

a protected connection,

FIG 4

a data packet,

FIG 5

another protected connection,

FIG 6

another data package,

FIGS. 7 to 11

Timing diagrams, and

FIG 12

a control system.

According to FIG 1 a control system has a plurality of sensors 1, a plurality of actuators 2 and a common central unit 3. The central unit 3 can be, for example, a programmable logic controller or a central unit of a programmable logic controller. The sensors 1 and the actuators 2 are peripheral devices.

The sensors 1 cyclically detect actual states Z of an industrial technical process 4, for example a production machine or a chemical process. The actual states Z can, for example, be position signals, temperatures, binary signals such as the response of a limit switch, etc. It is assumed below that each sensor 1 detects an individual actual state Z. The sensors 1 could, however, also detect several actual states Z in each case.

The sensors 1 and the central unit 3 are connected to one another via a first open communication network 5. Within the first open communication network 5, the sensors 1 transmit the actual states Z recorded by them to the central unit 3 via a first protected connection 6. The first protected connection 6 is common to the sensors 1.

The communication network 5 is an open communication network because other components 7 can also be connected to the communication network 5, which independently transmit data different from the actual states Z via the communication network 5 can transfer. It is possible that the other components 7 only communicate with one another, but neither with the central unit 3 nor with the sensors 1. However, it is also possible that the other components 7 also communicate with the central unit 3 and / or with the sensors 1. In this case, however, the communication takes place outside of the control method according to the invention. The first open communication network 5 can be designed as an Ethernet, for example.

The first protected connection 6 is a protected connection because the type of connection 6 ensures that data transmitted from the sensors 1 via the first protected connection 6 to the central unit 3 within a predetermined maximum latency period at the central unit 3 arrive. Between the feeding of the data - here the respective actual state Z detected by the respective sensor 1 - into the first open communication network 5 by one of the sensors 1 and the arrival of this data at the central unit 3, the maximum latency period elapses. The latency period can differ from sensor 1 to sensor 1. But it is given for each sensor 1. An example of a protected connection 6 is a stream as it is defined, for example, by the AVB (= Audio / Video Bridging) Task Group and in particular by the TSN (= Time-Sensitive Networking) Task Group in the international standard IEEE 802.1.

A stream is usually defined between a single sender (talker) and a single receiver (listener). In the present case, however, there are several transmitters, namely the sensors 1. However, it is known to set up a stream between several transmitters and a single receiver (here the central unit 3). Purely by way of example, the PCT / EP2017 / 055643 , submitted on March 10, 2017, applicant Siemens AG. More details on setting up will be explained later.

The transmission of the actual states Z to the central unit 3, like the detection of the actual states Z by the sensors 1, takes place cyclically. Thus, transmitted within a predetermined time window T in accordance with the representation in FIG FIG 2 each sensor 1 sends the actual states Z recorded by it once (1x) to the central unit 3. The totality of the actual states Z forms the so-called process image of the PAE inputs.

The central unit 3 cyclically ascertains control signals C for the industrial technical process 4 in a manner known per se. The central unit 3 takes into account the actual states Z transmitted to it and possibly other internal data such as flags and timers. The corresponding procedure is generally known for programmable logic controllers. The control signals C can, for example, cause a heater, a servomotor, a lamp, etc. to be switched on or off. It can be a binary signal, a discrete signal or an analog signal. The latter are determined in digital form by the central unit 3 and transmitted to the actuators 2. They are only converted from digital to analog form after they have been transmitted to the actuators 2.

The central unit 3 and the actuators 2 are - analogously to the connection of the central unit 3 and the sensors 1 - connected to one another via a second open communication network 8. Within the second open communication network 8, the central unit 3 transmits the control signals C determined by it to the actuators 2 via a second protected connection 9. The second protected connection 9 is common to the actuators 2. The entirety of the control signals C forms the so-called process image of the outputs PAA. It is assumed below that the central unit 3 determines an individual control signal C for each actuator 2. The central unit 3 could, however, also determine a plurality of control signals C for each of the actuators 2.

The communication network 8 is an open communication network because other components 10 can also be connected to the communication network 8, which components can independently transmit data different from the control signals C via the second open communication network 8. It is possible that the other components 10 only communicate with one another, but neither with the central unit 3 nor with the actuators 2. However, it is also possible that the other components 10 also communicate with the central unit 3 and / or with the actuators 2. In this case, however, the communication takes place outside of the control method according to the invention. The second open communication network 8 can be designed as an Ethernet, for example, analogously to the communication network 5.

The second protected connection 9 is a protected connection because the type of connection 9 ensures that data transmitted from the central unit 3 to the actuators 2 via the second protected connection 9 within a predetermined maximum latency period for the actuators 2 arrive. Between the feeding of the data - here the respective control signal C intended for the respective actuator 2 - into the second open communication network 8 by the central unit 3 and the arrival of this data at the respective actuator 2, the maximum latency period elapses. The latency period can differ from actuator 2 to actuator 2. However, it is given for each actuator 2.

An example of a protected connection 9 is - as before - a stream as defined by the Audio / Video Bridging Task Group and in particular by the Time-Sensitive Networking Task Group in the international standard IEEE 802.1. Analogously to setting up a stream between a number of transmitters and a single receiver, it is also known to set up a stream between a single transmitter (here the central unit 3) and a number of receivers (here the actuators 2). This corresponds to the standard principle as defined by the AVB Task Group and in particular by the TSN Task Group in the IEEE 802.1 standard.

The transmission of the control signals C to the actuators 2 takes place cyclically. Thus transmitted within the predetermined time window T in accordance with the representation in FIG FIG 2 the central unit 3 sends the control signal C intended for the respective actuator 2 once (1x) to each actuator 2. The actuators 2 then act on the industrial technical process 4 in accordance with the control signals C transmitted to them. The action on the industrial technical process 4 also takes place cyclically.

The establishment of the second protected connection 9 is explained below. It is assumed here that the logical or physical structure of the second open communication network 8, as far as the central unit 3 and the actuators 2 are concerned, as in FIG FIG 3 is shown. However, the topology of the second open communication network 8 can also be different. In particular, it does not have to be known in advance.

According to FIG 3 each actuator 2 is connected to the central unit 3 via a respective sequence of line sections 11 of the second open communication network 8. The communication from the central unit 3 to the respective actuator 2 takes place via the respective sequence. Line sections 11 directly adjoining one another are connected to one another via a respective node 12. The reference numerals 11 and 12 are in FIG FIG 3 only shown for the path from the central unit 3 to one of the actuators 2.

The nodes 12 - usually referred to as bridges in the case of a stream - forward the data received from them. In the case of an unprotected connection, the delay that occurs is not predetermined. In the case of a protected connection - here the second protected connection 9 - the nodes 12 forward the data they have received with a respective maximum delay time. The nodes 12 can ensure this because when setting up of the second protected connection 9, each node 12 involved checks whether its internal resources are sufficient for the performance required within the framework of the protected connection to be set up (in particular with regard to data volume and data throughput). If this is the case, the respective node 12 reserves these resources for the protected connection to be set up. Otherwise, a corresponding message is sent to a facility setting up the protected connection. The protected connection is not established in this case. On the basis of this procedure, each node 12 involved can guarantee that the required performance will be maintained during later operation.

During later operation, it is true that it is not known in advance with which actual delay time the respective node 12 forwards the control signals C fed into the second protected connection 9 by the central unit 3. However, the maximum delay time (i.e. its maximum value) remains guaranteed. In individual cases, it may still be possible that the respective node 12 can also specify a minimum delay time. Alternatively, it can be assumed that the minimum delay time (that is, its minimum value) has the value zero.

The data transmission via the second protected connection 9 takes place in the form of individual data packets (frames), which according to FIG 4 a header 13 and payload 14 include. The control signals C are components of the user data 14. The header 13 comprises, on the one hand, an identifier K for the second protected connection 9 as such and, on the other hand, a logical source address SA and a logical destination address DA. In order to set up the second protected connection 9, it is therefore particularly necessary for each actuator 2 to be assigned one and the same logical destination address DA. As a result, the central unit 2 transmits all control signals C to each actuator 2 in the second protected connection 9. This means that each actuator 2 is able to filter out and use the control signals C relevant to it. Temporal conflicts between the individual control signals C cannot occur with this procedure, since there is only a single transmitter (the central unit 3) and the central unit 3 is consequently the root of the second protected connection 9.

The establishment of the first protected connection 6 will now be explained. It is assumed here that the logical or physical structure of the first open communication network 5, as far as the central unit 3 and the sensors 1 are concerned, as in FIG FIG 5 is shown. However, the topology of the first open communication network 5 can also be different. In particular, it does not have to be known in advance.

According to the representation in FIG 5 the topology of the first open communication network 5 is similar to the topology of the second open communication network 8. However, this is not absolutely necessary.

According to FIG 5 each sensor 1 is connected to the central unit 3 via a sequence of line sections 15 of the first open communication network 5. The communication from the sensors 1 to the central unit 3 takes place via the respective sequence. Directly adjacent line sections 15 are each connected to one another via a node 16. Numbers 15 and 16 are in FIG FIG 5 only shown for the path from one of the sensors 1 to the central unit 2.

The nodes 16 - usually referred to as bridges in a stream - forward the data received from them. In the case of an unprotected connection, the delay that occurs is not predetermined. In the case of a protected connection - here the first protected connection 6 - the nodes 16 forward the data they have received with a respective maximum delay time. The nodes 16 can ensure this because when setting up of the first protected connection 6, each node 16 involved checks whether its internal resources are sufficient for the performance required in the context of the protected connection to be set up (in particular with regard to data volume and data throughput). If this is the case, the respective node 16 reserves these resources. Otherwise, a corresponding message is sent to a facility setting up the protected connection. The protected connection is not established in this case. On the basis of this procedure, each node 16 involved can ensure that the required performance is maintained during later operation.

During later operation it is not known in advance with which actual delay time the respective node 16 forwards the actual states Z fed into the first protected connection 6 by the respective sensor 1. However, the maximum delay time (i.e. its maximum value) remains guaranteed. In individual cases, it may still be possible for the respective node 16 to also specify a minimum delay time. Alternatively, it can be assumed that the minimum delay time (that is, its minimum value) has the value zero.

The data transmission via the first protected connection 6 takes place in the same way as via the second protected connection 9. The transmission thus takes place in the form of individual data packets (frames), which according to FIG FIG 6 a header 17 and payload data 18. The actual states Z are components of the useful data 18. The header 17 comprises, on the one hand, an identifier K 'for the first protected connection 6 as such and, on the other hand, a logical source address SA' and a logical destination address DA '. To set up the first protected connection 6, it is possible for each sensor 1 to be assigned one and the same logical source address SA '. However, it is also possible for the sensors 1 to be assigned different logical source addresses SA '. In both cases the sensors 1 in the first protected connection 6 transmit all the actual states Z to the central unit 3.

In contrast to the second protected connection 9, there are several transmitters in the first protected connection 6, namely the sensors 1. However, the sensors 1 “know” nothing of one another. If the sensors 1 are therefore only given to transmit their respective actual states Z to the central unit 3 via the first protected connection 6 within the time window T, it can happen that actual states Z, which are transmitted from various sensors 1 to the central unit 3, arrive at the central unit 3 or one of the nodes 16 at the same time, so that they cannot be received by the central unit 3 or the corresponding node 16. In this case, data would be lost.

In order to avoid such temporal conflicts when transmitting the actual states Z to the central unit 3, as shown in FIG FIG 7 Within the time window T, each sensor 1 is assigned a respective sub-area 19 on the transmitter side. The sensor 1 feeds the actual states Z detected by it to the first open communication network 5 within the sub-area 19 assigned to it. The respective sub-area 19 does not relate to the entire transmission from the respective sensor 1 to the central unit 3, but only to the first link in the transmission chain, ie feeding into the line section 15 adjacent to the respective sensor 1. The possibility of assigning Transmitter-side sub-areas 19 to a specific transmitter (here the sensors 1) is known as such. Purely by way of example, reference can again be made to the international standard IEEE 802.1 and in particular to the extensions in IEEE 802.1 Qcc. Corresponding procedures are also known from PROFINET IRT.

The receiver-side partial areas 20 correspond to the transmitter-side partial areas 19. The receiver-side partial areas 20 are those time ranges within which the actual states Z fed into the first open communication network 5 by one of the sensors 1 in its sub-area 19 on the transmitter side arrive at the central unit 3. The receiver-side partial areas 20 must be disjoint from one another. In order to ensure this, the respective sub-area 19 on the transmitter side is initially only provisionally set for each sensor 1. For this provisional approach - but only for this one - the subareas 19 on the transmitter side can even be used as shown in FIG FIG 8 to match. It is then determined for each transmitter-side sub-area 19 which receiver-side sub-area 20 on the central unit 3 side corresponds to the respective transmitter-side sub-area 19. FIG 8 also shows the corresponding partial areas 20 on the receiver side. The partial areas 20 on the receiver side are shown in FIG FIG 8 displayed at different heights, despite their temporal overlap in FIG 8 to be able to recognize them as such and to be able to distinguish them from one another.

It is now crucial that according to FIG 9 for each transmitter-side sub-area 19, the time offset A to the corresponding receiver-side sub-area 20 is constant. So if according to the dashed representation in FIG 9 a sender-side sub-area 19 is shifted by a certain time period δt, the corresponding receiver-side sub-area 20 is also shifted by precisely this time period δt. The time offset A is therefore retained unchanged. It is therefore possible, as shown in FIG 8 first to determine the associated receiver-side partial areas 20 for arbitrarily applied transmitter-side partial areas 19 and then the receiver-side partial areas 20 according to the illustration in FIG FIG 10 to be arranged within the time window T in such a way that they are disjoint from one another. A safety distance is preferably maintained between the partial areas 20 on the receiver side. However, this is not absolutely necessary.

The associated time shifts can easily be determined. Due to the fact that the time offset A with respect to the respective corresponding transmitter-side partial area 19 does not change, the corresponding transmitter-side partial areas 19 can thus also be determined without further ado. It is only necessary to shift the partial areas 19 on the transmitter side 1: 1 for the temporal shift of the respective partial area 20 on the receiver side.

It is possible that the transmitter-side partial areas 19 overlap in time after the shift. However, this is not critical, since it does not depend on a possible overlap of the partial areas 19 on the transmitter side, but only on avoiding an overlap of the partial areas 20 on the receiver side.

In order to implement the procedure explained above, the receiver-side partial areas 20 must be known. The determination of the receiver-side subarea 20 is explained below for an individual transmitter-side subarea 19. The procedure is, however, valid in a completely analogous manner for all sub-areas 19 on the transmitter side.

In the following, according to the representation in FIG 11 - purely by way of example - it is assumed that the corresponding sensor 1 is connected to the central unit 3 via four sequentially successive line sections 15 and, correspondingly, via three nodes 16 arranged between each two of these four line sections 15. It is also assumed that the transmission times via the line sections 15 as such can be neglected. If necessary, however, such transmission times can also be taken into account.

The sensor 1 under consideration feeds the actual states Z recorded by it to the first open communication network 5 within the time window T in the sub-area 19 assigned to it.

The boundaries of this sub-area 19 are given the reference symbols t1 and t2 below. The difference between the two limits, that is to say t2-t1, corresponds to the duration of the corresponding sub-area 19 on the transmitter side.

Due to the neglect of the transmission times via the line sections 15 as such, the transmitted actual states Z arrive at the node 16 adjacent to the sensor 1 in the same time period - that is, in the interval extending from t1 to t2. This node 16 feeds the actual states Z to the next line section 15 after a delay time. The exact value of the delay time is not known. It is known, however, that the delay time has a minimum value T1 and a maximum value T1 ′.

The earliest possible point in time at which this node 16 feeds the actual states Z to the next line section 15 is accordingly t1 + T1. The latest possible point in time at which this node 16 ends the supply of the actual states Z to the next line section 15 is accordingly still t2 + T1 '.

Due to the neglect of the transmission times via the line sections 15 as such, the transmitted actual states Z arrive at the next node 16 in the same time period - that is, in the interval extending from t1 + T1 to t2 + T1 '. This node 16 also feeds the actual states Z to the next line section 15 as seen from it after a delay time. The exact value of the delay time is again not known. It is known, however, that the delay time has a minimum value T2 and a maximum value T2 '.

The earliest possible point in time at which this node 16 feeds the actual states Z to the next line section 15 is accordingly t1 + T1 + T2. The latest possible point in time at which this node 16 starts supplying the actual states Z ended to the next line section 15, is accordingly still at t2 + T1 '+ T2'.

This procedure can be repeated for each node 16. In the example given, in which three nodes 16 are arranged between the sensor 1 and the central unit 3, the earliest possible point in time at which the actual states Z of the sensor 1 under consideration arrive at the central unit 3 is t1 + T1 + T2 + T3, where T3 is the minimum delay time of the node 16 adjacent to the central unit 3. In an analogous manner, the latest possible point in time at which the transmission of the actual states Z of the sensor 1 under consideration to the central unit 3 has ended is t1 + T1 '+ T2' + T3 ', where T3' is the maximum delay time of the to the central unit 3 adjacent node 16 is.

As a result, for each sensor 1 and its transmitter-side sub-area 19, the respective receiver-side sub-area 20, starting from the transmitter-side sub-area 19 assigned to the respective sensor 1, can be determined by adding up the predetermined minimum values T1, T2, T3 etc. of the nodes 16 and adding the determine predetermined maximum values T1 ', T2', T3 'etc. of the nodes 16.

The sender-side partial areas 19 are determined automatically using a P2P protocol, for example on the basis of LRP (= link local reservation protocol). After the partial areas 19 on the transmitter side have been determined, these are transmitted to the individual sensors 1, for example via the first open communication network 5, but outside the first protected connection 6.

The number of data packets that are transmitted per time window T via the first open communication network 5 and here via the first protected connection 6 from the sensors 1 to the central unit 3 is generally as large as the number of sensors 1. This is therefore the case Case, because each sensor 1 the actual states Z recorded by it are transmitted to the central unit 3 via a separate data packet. The resources of the participating nodes 16 of the first open communication network 5 must therefore be determined in the context of the establishment of the first protected connection 6 in such a way that they within the specified maximum delay time Ti '(with i = 1, 2, 3 etc.) the can forward the corresponding number of data packets.

At least the node 16 of the first open communication network 5 directly adjacent to the central unit 3 must generally be able to handle all the data packets transmitted by the sensors 1, that is, regardless of which sensor 1 the corresponding data packet originates from. With regard to the other nodes 16 of the first open communication network 5, it may be sufficient to reserve their resources to a reduced extent for the first protected connection 6.

In order for the sensors 1 to adhere to the transmitter-side subregions 19, it is also necessary for the sensors 1 to be synchronized with one another. However, the synchronization as such is not the subject of the present invention. Synchronization options are also well known to those skilled in the art. They therefore do not have to be explained in detail at this point.

As far as explained above, the first open communication network 5 and the second open communication network 8 are communication networks that are different from one another. However, this is not absolutely necessary. Rather, the first open communication network 5 and the second open communication network 8 can be used as shown in FIG FIG 12 have at least one common line section 11, 15 via which both the transmission of the actual states Z to the central unit 3 and the transmission of the control signals C to the actuators 2 take place. In particular, the central unit 1 is usually only connected to the two via a single line section 11, 15 open communication networks 5, 8 connected. Further line sections 11, 15 can also be part of both open communication networks 5, 18. Furthermore, it is also possible for peripheral units to be present which include both the functionality of a sensor 1 and the functionality of an actuator 2 (so-called mixed I / O units). In this case, the transmission of both the actual states Z to the central unit 3 and the control signals C from the central unit 3 via the same line sections 11, 15 occurs completely automatically.

In summary, the present invention relates to the following facts:
Sensors 1 of a control system cyclically detect actual states Z of an industrial technical process 4 and transmit them via a first protected connection 6 of a first open communication network 5 common to sensors 1 to a common central unit 3 of the control system. As a result, within a predetermined time window T, each sensor 1 transmits the actual states Z it detects once to the central unit 3. The central unit 3 cyclically determines control signals C for the industrial technical process 4, taking into account the actual states Z transmitted to it, and transmits them via a number of actuators 2 of the control system common second protected connection 9 of a second open communication network 8 to the actuators 2. The central unit 3 thereby transmits within the predetermined time window T to each actuator 2 once the control signals C intended for the respective actuator 2. The actuators 2 act cyclically according to the control signals C transmitted to them on the industrial technical process 4. Each sensor 1 is assigned a respective transmitter-side sub-area 19 within the time window T, within which the respective sensor 19 feeds the actual states Z detected by it to the first open communication network 5. The transmitted actual states Z occur within a sub-area 19 corresponding to the respective transmitter-side respective receiver-side sub-area 20 of the time window T at the central unit 3. The transmitter-side partial areas 19 of the sensors 1 are determined in such a way that the receiver-side partial areas 20 are disjoint from one another.

The present invention has many advantages. In particular, there is efficient and reliable and also deterministic communication between the sensors 1, the central unit 3 and the actuators 2 of the control system. If the communication networks 5, 8 have a realistic depth of up to seven nodes 12, 16, there are usually delay times between the supply of actual states Z or control signals C to the communication network 5, 8 and their arrival at the central unit 3 or the actuators 2 of a few milliseconds (usually a maximum of 5 ms). Resources in the nodes 12, 16 can be saved. In particular, only one entry for the two protected connections 6, 9 has to be managed in the reservation protocol and the data plane of the first and second open communication networks 5, 8. In the event of an error, the diagnosis is simplified since in both communication directions - towards the central unit 3 and away from the central unit 3 - only a single protected connection 6, 9 has to be checked. Knowledge of the topology of the communication networks 5, 8 is not required. A corresponding reservation of resources of the corresponding nodes 16 is only required in those nodes 16 via which the actual states Z of several sensors 1 are forwarded.

Although the invention has been illustrated and described in more detail by the preferred exemplary embodiment, the invention is not restricted by the disclosed examples and other variations can be derived therefrom by the person skilled in the art without departing from the scope of protection of the invention.

Claims (6)

1. Control method for a technical industrial process (4),

   - wherein sensors (1) of a control system cyclically detect actual states (Z) of the technical industrial process (4) and transmit same to a common central unit (3) of the control system via a first protected connection (6) of a first open communication network (5), said connection being common to the sensors (1) and each sensor (1) transmitting the actual states (Z) detected by the sensor to the common central unit (3) once within a specified time window (T),

   - wherein the common central unit (3) cyclically ascertains control signals (C) for the technical industrial process (4) while taking into consideration the actual states (Z) transmitted to the central unit and transmits said control signals to multiple actuators (2) of the control system via a second protected connection (9) of a second open communication network (8), said connection being common to the actuators (2), the common central unit (3) thereby transmitting the control signals (C) determined for the respective actuator (2) to each actuator (2) once within the specified time window (T),

   - wherein the actuators (2) act cyclically on the technical industrial process (4) in a manner corresponding to the control signals (C) transmitted to the actuators,

   - wherein each sensor (1) is assigned a respective transmitter-side sub-region (19) within which the respective sensor (19) supplies the actual states (Z) detected by said sensor to the first open communication network (5) within the time window (T),

   - wherein the transmitted actual states (Z) arrive in the central unit (3) within a respective receiver-side sub-region (20) of the time window (T), said sub-region corresponding to the respective transmitter-side sub-region (19), and

   - wherein the transmitter-side sub-regions (19) of the sensors (1) are determined such that the receiver-side sub-regions (20) are disjointed relative to one another.
2. Control method according to claim 1,
   characterised in that

   - to ascertain the transmitter-side sub-regions (19) for the sensors (1), the respective transmitter-side sub-region (19) is first temporarily set up within the time window (T),

   - it is then determined for each transmitter-side sub-region (19) which receiver-side sub-region (20) corresponds to the respective transmitter-side sub-region (19) on the side of the common central unit (3), and

   - finally, the transmitter-side sub-regions (19) are shifted within the time window (T) such that the receiver-side sub-regions (20) are disjointed relative to one another.
3. Control method according to claim 2,
   characterised in that

   - communication from the respective sensor (1) to the common central unit (3) takes place via a respective sequence of sequentially successive line sections (15) of the first open communication network (5),

   - directly adjacent line sections (15) of the respective sequence are connected to one another via one node (16) in each case,

   - each node (16) forwards the actual states (Z) transmitted by the respective sensor (1) with a respective delay time,

   - the delay time of the respective node (16) lies between a minimum value (T1, T2, T3) predetermined for the respective node (16) and a maximum value (T1', T2', T3') predetermined for the respective node (16) and

   - the respective receiver-side sub-region (20), starting from the transmitter-side sub-region (19) assigned to the respective sensor (1), is ascertained by adding up the predetermined minimum values (T1, T2, T3) of the nodes (16) and adding up the predetermined maximum values (T1', T2', T3') of the nodes (16).
4. Control method according to claim 3,
   characterised in that the predetermined minimum values (T1, T2, T3) are set to zero.
5. Control method according to one of the above claims,
   characterised in that the first open communication network (5) and the second open communication network (8) have at least one common line section (11, 15) via which both the transmission of the actual states (Z) to the common central unit (3) and the transmission of the control signals (C) to the actuators (2) takes place.
6. Control system for a technical industrial process,

   - wherein the control system has a plurality of sensors (1), a plurality of actuators (2) and a common central unit (3),

   - wherein the sensors (1) and the common central unit (3) are connected to one another via a first open communication network (5),

   - wherein the common central unit (3) and the actuators (2) are connected to one another via a second open communication network (8),

   - wherein the sensors (1), the common central unit (3), the actuators (2), the first open communication network (5) and the second open communication network (8) are designed in such a way that during operation they interact according to a control method according to one of the above claims.

Images